Journal of Cardiovascular Nursing
Vol. 30, No. 4S, pp S44YS51 x Copyright B 2015 Wolters Kluwer Health, Inc. All rights reserved.
Care in the Home for the Management of Chronic Heart Failure Systematic Review and Cost-Effectiveness Analysis Jennifer Fergenbaum, PhD; Sarah Bermingham, MSc; Murray Krahn, MD; David Alter, MD; Catherine Demers, MD Background: The objective of this study was to determine the effect of care in the home (CHM) compared with usual care (UC) in patients with chronic heart failure (CHF) on clinical outcomes and healthcare use including a cost-effectiveness analysis. Methods: A systematic literature search on MEDLINE, EMBASE, Cumulative Index to Nursing and Allied Health Literature, the Cochrane Library, as well as Centre for Reviews and Dissemination was conducted to identify randomized controlled trials comparing CHM with UC in CHF. The randomized controlled trials meeting inclusion criteria were meta-analyzed by outcome, and the quality of evidence for each outcome was evaluated using Grading of Recommendations Assessment, Development, and Evaluation system. A cost-effectiveness model was developed to estimate costs and quality-adjusted life years. Results: Six randomized controlled trials were identified from 1277 citations. Care in the home was predominately provided by a single health professional consisting of nurse-led education of varying duration and frequency. One study included pharmacist-led CHM. Care in the home showed a decreased risk for all-cause mortality and hospitalizations combined (risk ratio, 0.88; 95% confidence interval [CI], 0.80Y0.97), but not all-cause mortality alone (risk ratio, 0.92; 95% CI, 0.81Y1.04). Care in the home resulted in fewer hospitalizations (mean difference, j1.03; 95% CI, j1.53 to j0.53) and fewer emergency department visits (mean difference, j1.32; 95% CI, j1.87 to j0.77). Quality of life also improved with CHM delivered by nurses. Critical appraisal of the quality of evidence suggests uncertainty in the estimates for a number of outcomes. Care in the home resulted in a savings of $10,665 and a gain of 0.11 quality-adjusted life years compared with UC. Conclusions: In conclusion, the beneficial effect of CHM in CHF is by reducing mortality and hospitalizations combined. Care in the home in CHF seems to be more effective and less costly compared with UC. KEY WORDS:
chronic heart failure, cost-effectiveness, meta-analysis, systematic review
C
hronic heart failure (CHF) is a complex syndrome describing a range of cardiac abnormalities, which leads to functional limitations and a number of symptoms including fatigue and dyspnea.1 Worldwide CHF affects 23 million individuals, and although its incidence may no longer be on the rise, its prevalence is increasing owing to improved disease prognosis.2 With an aging population, care of patients with CHF continues to be at the forefront of chronic disease management.3
There is an ongoing need to optimize CHF management to meet patients’ challenging needs and the demand to provide complex interventions simultaneously.4 The most effective care models promote patient self-care with a focus on self-care maintenance (eg, medication adherence) and self-care management (eg, symptom recognition and therapeutic evaluation).5 Care in the home (CHM) has the potential to be a critical component of CHF management because it includes a variety Catherine Demers, MD
Jennifer Fergenbaum, PhD Clinical Epidemiologist, Health Quality Ontario, Evidence Development and Standards Branch, Toronto, Canada.
Sarah Bermingham, MSc Health Economist, Health Quality Ontario, Evidence Development and Standards Branch, Toronto, Canada; and The Toronto Health Economic and Technology Assessment (THETA) Collaborative, Ontario, Canada.
Murray Krahn, MD Director, The Toronto Health Economic and Technology Assessment (THETA) Collaborative, Ontario, Canada.
David Alter, MD Senior Scientist and Cardiologist, Department of Medicine, University of Toronto, Cardiologist, Toronto Rehabilitation Institute-University Health Network, Institute for Clinical Evaluative Sciences, Ontario, Canada.
Cardiologist, Division of Cardiology, Department of Medicine, McMaster University, Hamilton General Hospital, Ontario, Canada. Health Quality Ontario is sponsored by the Ontario Ministry of Health and Long-Term Care. The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of Health Quality Ontario. The authors have no conflicts of interest to disclose. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.jcnjournal.com).
Correspondence Jennifer Fergenbaum, PhD, Health Quality Ontario, Evidence Development and Standards Branch, 130 Bloor St W, 10th Flr, Toronto, Ontario, M5S 1N5 ([email protected]). DOI: 10.1097/JCN.0000000000000235
S44 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Care in the Home in Chronic Heart Failure Management S45
of home services such as health promotion and teaching, clinical care, end-of-life care, rehabilitation, social adaptation and integration, and support for informal caregivers.6 For CHF management in the home, the aim is to improve health and prevent the need for long-term care or immediate care.7 Care in the home has received little attention as an intervention in trials of CHF multidisciplinary care. Investigations have identified a trend toward decreased mortality, readmissions, and costs.8Y10 However, the interpretation of this literature is challenging because these studies included interventions or comparators that were heterogeneous. Further, what is lacking is the consideration that, although all patients with CHF may receive outpatient clinic-based multidisciplinary care as part of postdischarge community-based care, not all patients with CHF receive CHM. Overall, the effect of CHM has been difficult to decipher. Therefore, the objectives of this study were to (1) conduct a systematic review and metaanalysis to determine the clinical effectiveness of CHM compared with usual care (UC) in patients with CHF and (2) to estimate the cost-effectiveness of CHM compared with UC in patients with CHF.
studies that (1) evaluated care provided as part of an outpatient clinic-based multidisciplinary program or selfmanagement program in the absence of a home visit by a health professional; (2) were part of a telemonitoring or telemedicine program; (3) were entirely a telephone-based home service; (4) consisted of hospice care, end-of-life care, or palliative care; (5) included different delivery models of home care; (6) were based on a model of transitional care, which describes a supported care pathway including the transition of patients from the hospital to home, which is initiated in the hospital; (7) examined early supported discharge; (8) were based on a model of hospital-at-home; and (9) focused on CHF-specific rehabilitation. Outcomes of interest included all-cause mortality and hospitalizations combined, all-cause mortality, cardiovascular-specific mortality, unplanned hospitalizations, CHF specific hospitalization, emergency department visits, as well as healthrelated quality of life. Animal or in vitro studies, abstracts, posters, case reports, letters or editorials, foreign language publications, and unpublished studies were also excluded. Reference lists were examined for additional relevant studies. Study Selection and Critical Appraisal
First, a systematic review and meta-analysis were performed to determine the clinical effectiveness of CHM compared with UC. Second, this information was then used to help inform the cost-effectiveness analysis (See Document, Supplemental Digital Content 1, for supplemental methods, http://links.lww.com/JCN/A9).
Full-text articles of studies conducted in CHF populations, with the defined intervention and comparator and that reported 1 outcome of interest or more, were retrieved. The quality of the clinical evidence was critically appraised using the Grading of Recommendations Assessment, Development, and Evaluation system, which evaluates the risk for bias, inconsistency, indirectness, and imprecision for each outcome.11
Literature Search
Analysis
OVID MEDLINE, EMBASE, Cumulative Index to Nursing and Allied Health Literature, the Cochrane Library, and the Centre for Reviews and Dissemination were searched from January 1, 2006 to January 25, 2012 to identify recent randomized controlled trials comparing CHM with UC in adults with CHF (Q18 years). The literature search was part of a larger initiative aimed to examine multiple chronic diseases (See Document, Supplemental Digital Content 1, for supplemental methods, http://links.lww.com/JCN/A9).
For each included randomized controlled trial the relevant aspects of study design, study population, and intervention characteristics were abstracted. A high level of education was defined as having completed a high school education or greater. The proportion of individuals having completed a high school education or greater was calculated as the average across the CHM and UC groups. When this information was missing but information on the proportion of primary schooling (or primary schooling or less) was available, the proportion of the population receiving a high level of education was calculated as the average proportion of primary schooling (or less) across the CHM and UC groups minus 100%. A meta-analysis was used to combine the results of studies by outcome using Review Manager version 5 software.12 Dichotomous outcomes were pooled using a fixed-effect (Mantel-Haenszel) model, and risk ratios with corresponding confidence intervals (CIs) were reported. Continuous outcomes were analyzed using an inverse variance method for pooling weighted mean differences, and mean differences with corresponding confidence intervals were reported. Statistical heterogeneity
Materials and Methods
Assessment of Study Eligibility Potentially relevant studies were identified on the basis of title and abstract. Potentially eligible studies were reviewed by a single reviewer. The inclusion criteria were studies in which (1) CHM was defined as at least 1 home visit by any type of health professional with the aim to improve the health and well-being of community-based patients with CHF, (2) included patients with CHF, and (3) studies that were randomized controlled trials of CHM versus UC (ie, no CHM). The exclusion criteria were
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
S46 Journal of Cardiovascular Nursing x July/August 2015 was assessed by considering the I2, a measure of consistency, and its corresponding # 2 test for significance, where P e .1 was used to indicate significant heterogeneity. When significant heterogeneity was observed, either a random-effects model or a subgroup analysis was considered. The cost-effectiveness of CHM compared with UC was estimated using the results obtained from the metaanalysis and health administrative databases from the Institute of Clinical and Evaluative Sciences (Toronto, Ontario, Canada). A phase-based method previously described by Wijeysundera et al10 was used to calculate real-world resource use, costs, and mortality per patient with CHF over 5 years. The effect of CHM on hospitalizations, emergency department visits, and all-cause mortality taken from the meta-analysis was applied to baseline cost and survival to estimate cost-effectiveness associated with CHM. Cost-effectiveness was expressed in terms of direct health care costs (2012 Canadian dollars) and quality-adjusted life years discounted at the standard annual rate of 5%.
Results A total of 1277 citations were identified by the literature search, of which 76 studies had the potential to be included and therefore underwent full text review. Three additional citations were identified through manual
searches of reference lists and databases. Overall, 6 randomized controlled trials met the inclusion criteria13Y18 (Figure 1). The study design and population characteristics of the included studies are shown in Table 1. Three of the included randomized controlled trials were conducted in Spain13,14,17 as well as 1 randomized controlled trials in each of the United Kingdom,16 Thailand,15 and Australia.18 Of the 6 randomized controlled trials, the study conducted in Thailand did not have suitable data for the metaanalysis.15 Therefore, only 5 randomized controlled trials comprised the pool of available data for the metaanalysis.13,14,16Y18 The severity of CHF ranged from class I (46%)13,15 and class I/II (approximately 46%)18 to class III/IV (approximately 34%)16 and class IV (51%),14 according to the New York Heart Association Functional Classification. One study did not provide information regarding severity.17 Small sample sizes predominated across the studies (
Vol. 30, No. 4S, pp S44YS51 x Copyright B 2015 Wolters Kluwer Health, Inc. All rights reserved.
Care in the Home for the Management of Chronic Heart Failure Systematic Review and Cost-Effectiveness Analysis Jennifer Fergenbaum, PhD; Sarah Bermingham, MSc; Murray Krahn, MD; David Alter, MD; Catherine Demers, MD Background: The objective of this study was to determine the effect of care in the home (CHM) compared with usual care (UC) in patients with chronic heart failure (CHF) on clinical outcomes and healthcare use including a cost-effectiveness analysis. Methods: A systematic literature search on MEDLINE, EMBASE, Cumulative Index to Nursing and Allied Health Literature, the Cochrane Library, as well as Centre for Reviews and Dissemination was conducted to identify randomized controlled trials comparing CHM with UC in CHF. The randomized controlled trials meeting inclusion criteria were meta-analyzed by outcome, and the quality of evidence for each outcome was evaluated using Grading of Recommendations Assessment, Development, and Evaluation system. A cost-effectiveness model was developed to estimate costs and quality-adjusted life years. Results: Six randomized controlled trials were identified from 1277 citations. Care in the home was predominately provided by a single health professional consisting of nurse-led education of varying duration and frequency. One study included pharmacist-led CHM. Care in the home showed a decreased risk for all-cause mortality and hospitalizations combined (risk ratio, 0.88; 95% confidence interval [CI], 0.80Y0.97), but not all-cause mortality alone (risk ratio, 0.92; 95% CI, 0.81Y1.04). Care in the home resulted in fewer hospitalizations (mean difference, j1.03; 95% CI, j1.53 to j0.53) and fewer emergency department visits (mean difference, j1.32; 95% CI, j1.87 to j0.77). Quality of life also improved with CHM delivered by nurses. Critical appraisal of the quality of evidence suggests uncertainty in the estimates for a number of outcomes. Care in the home resulted in a savings of $10,665 and a gain of 0.11 quality-adjusted life years compared with UC. Conclusions: In conclusion, the beneficial effect of CHM in CHF is by reducing mortality and hospitalizations combined. Care in the home in CHF seems to be more effective and less costly compared with UC. KEY WORDS:
chronic heart failure, cost-effectiveness, meta-analysis, systematic review
C
hronic heart failure (CHF) is a complex syndrome describing a range of cardiac abnormalities, which leads to functional limitations and a number of symptoms including fatigue and dyspnea.1 Worldwide CHF affects 23 million individuals, and although its incidence may no longer be on the rise, its prevalence is increasing owing to improved disease prognosis.2 With an aging population, care of patients with CHF continues to be at the forefront of chronic disease management.3
There is an ongoing need to optimize CHF management to meet patients’ challenging needs and the demand to provide complex interventions simultaneously.4 The most effective care models promote patient self-care with a focus on self-care maintenance (eg, medication adherence) and self-care management (eg, symptom recognition and therapeutic evaluation).5 Care in the home (CHM) has the potential to be a critical component of CHF management because it includes a variety Catherine Demers, MD
Jennifer Fergenbaum, PhD Clinical Epidemiologist, Health Quality Ontario, Evidence Development and Standards Branch, Toronto, Canada.
Sarah Bermingham, MSc Health Economist, Health Quality Ontario, Evidence Development and Standards Branch, Toronto, Canada; and The Toronto Health Economic and Technology Assessment (THETA) Collaborative, Ontario, Canada.
Murray Krahn, MD Director, The Toronto Health Economic and Technology Assessment (THETA) Collaborative, Ontario, Canada.
David Alter, MD Senior Scientist and Cardiologist, Department of Medicine, University of Toronto, Cardiologist, Toronto Rehabilitation Institute-University Health Network, Institute for Clinical Evaluative Sciences, Ontario, Canada.
Cardiologist, Division of Cardiology, Department of Medicine, McMaster University, Hamilton General Hospital, Ontario, Canada. Health Quality Ontario is sponsored by the Ontario Ministry of Health and Long-Term Care. The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of Health Quality Ontario. The authors have no conflicts of interest to disclose. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.jcnjournal.com).
Correspondence Jennifer Fergenbaum, PhD, Health Quality Ontario, Evidence Development and Standards Branch, 130 Bloor St W, 10th Flr, Toronto, Ontario, M5S 1N5 ([email protected]). DOI: 10.1097/JCN.0000000000000235
S44 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Care in the Home in Chronic Heart Failure Management S45
of home services such as health promotion and teaching, clinical care, end-of-life care, rehabilitation, social adaptation and integration, and support for informal caregivers.6 For CHF management in the home, the aim is to improve health and prevent the need for long-term care or immediate care.7 Care in the home has received little attention as an intervention in trials of CHF multidisciplinary care. Investigations have identified a trend toward decreased mortality, readmissions, and costs.8Y10 However, the interpretation of this literature is challenging because these studies included interventions or comparators that were heterogeneous. Further, what is lacking is the consideration that, although all patients with CHF may receive outpatient clinic-based multidisciplinary care as part of postdischarge community-based care, not all patients with CHF receive CHM. Overall, the effect of CHM has been difficult to decipher. Therefore, the objectives of this study were to (1) conduct a systematic review and metaanalysis to determine the clinical effectiveness of CHM compared with usual care (UC) in patients with CHF and (2) to estimate the cost-effectiveness of CHM compared with UC in patients with CHF.
studies that (1) evaluated care provided as part of an outpatient clinic-based multidisciplinary program or selfmanagement program in the absence of a home visit by a health professional; (2) were part of a telemonitoring or telemedicine program; (3) were entirely a telephone-based home service; (4) consisted of hospice care, end-of-life care, or palliative care; (5) included different delivery models of home care; (6) were based on a model of transitional care, which describes a supported care pathway including the transition of patients from the hospital to home, which is initiated in the hospital; (7) examined early supported discharge; (8) were based on a model of hospital-at-home; and (9) focused on CHF-specific rehabilitation. Outcomes of interest included all-cause mortality and hospitalizations combined, all-cause mortality, cardiovascular-specific mortality, unplanned hospitalizations, CHF specific hospitalization, emergency department visits, as well as healthrelated quality of life. Animal or in vitro studies, abstracts, posters, case reports, letters or editorials, foreign language publications, and unpublished studies were also excluded. Reference lists were examined for additional relevant studies. Study Selection and Critical Appraisal
First, a systematic review and meta-analysis were performed to determine the clinical effectiveness of CHM compared with UC. Second, this information was then used to help inform the cost-effectiveness analysis (See Document, Supplemental Digital Content 1, for supplemental methods, http://links.lww.com/JCN/A9).
Full-text articles of studies conducted in CHF populations, with the defined intervention and comparator and that reported 1 outcome of interest or more, were retrieved. The quality of the clinical evidence was critically appraised using the Grading of Recommendations Assessment, Development, and Evaluation system, which evaluates the risk for bias, inconsistency, indirectness, and imprecision for each outcome.11
Literature Search
Analysis
OVID MEDLINE, EMBASE, Cumulative Index to Nursing and Allied Health Literature, the Cochrane Library, and the Centre for Reviews and Dissemination were searched from January 1, 2006 to January 25, 2012 to identify recent randomized controlled trials comparing CHM with UC in adults with CHF (Q18 years). The literature search was part of a larger initiative aimed to examine multiple chronic diseases (See Document, Supplemental Digital Content 1, for supplemental methods, http://links.lww.com/JCN/A9).
For each included randomized controlled trial the relevant aspects of study design, study population, and intervention characteristics were abstracted. A high level of education was defined as having completed a high school education or greater. The proportion of individuals having completed a high school education or greater was calculated as the average across the CHM and UC groups. When this information was missing but information on the proportion of primary schooling (or primary schooling or less) was available, the proportion of the population receiving a high level of education was calculated as the average proportion of primary schooling (or less) across the CHM and UC groups minus 100%. A meta-analysis was used to combine the results of studies by outcome using Review Manager version 5 software.12 Dichotomous outcomes were pooled using a fixed-effect (Mantel-Haenszel) model, and risk ratios with corresponding confidence intervals (CIs) were reported. Continuous outcomes were analyzed using an inverse variance method for pooling weighted mean differences, and mean differences with corresponding confidence intervals were reported. Statistical heterogeneity
Materials and Methods
Assessment of Study Eligibility Potentially relevant studies were identified on the basis of title and abstract. Potentially eligible studies were reviewed by a single reviewer. The inclusion criteria were studies in which (1) CHM was defined as at least 1 home visit by any type of health professional with the aim to improve the health and well-being of community-based patients with CHF, (2) included patients with CHF, and (3) studies that were randomized controlled trials of CHM versus UC (ie, no CHM). The exclusion criteria were
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
S46 Journal of Cardiovascular Nursing x July/August 2015 was assessed by considering the I2, a measure of consistency, and its corresponding # 2 test for significance, where P e .1 was used to indicate significant heterogeneity. When significant heterogeneity was observed, either a random-effects model or a subgroup analysis was considered. The cost-effectiveness of CHM compared with UC was estimated using the results obtained from the metaanalysis and health administrative databases from the Institute of Clinical and Evaluative Sciences (Toronto, Ontario, Canada). A phase-based method previously described by Wijeysundera et al10 was used to calculate real-world resource use, costs, and mortality per patient with CHF over 5 years. The effect of CHM on hospitalizations, emergency department visits, and all-cause mortality taken from the meta-analysis was applied to baseline cost and survival to estimate cost-effectiveness associated with CHM. Cost-effectiveness was expressed in terms of direct health care costs (2012 Canadian dollars) and quality-adjusted life years discounted at the standard annual rate of 5%.
Results A total of 1277 citations were identified by the literature search, of which 76 studies had the potential to be included and therefore underwent full text review. Three additional citations were identified through manual
searches of reference lists and databases. Overall, 6 randomized controlled trials met the inclusion criteria13Y18 (Figure 1). The study design and population characteristics of the included studies are shown in Table 1. Three of the included randomized controlled trials were conducted in Spain13,14,17 as well as 1 randomized controlled trials in each of the United Kingdom,16 Thailand,15 and Australia.18 Of the 6 randomized controlled trials, the study conducted in Thailand did not have suitable data for the metaanalysis.15 Therefore, only 5 randomized controlled trials comprised the pool of available data for the metaanalysis.13,14,16Y18 The severity of CHF ranged from class I (46%)13,15 and class I/II (approximately 46%)18 to class III/IV (approximately 34%)16 and class IV (51%),14 according to the New York Heart Association Functional Classification. One study did not provide information regarding severity.17 Small sample sizes predominated across the studies (
